Wikipedia talk:Article Feedback/Feedback response guidelines/Archive 1

Archive 1

The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

---

First Round of Discussion

Discuss

Marking a feedback as helpful/unhelpful

It is impossible to trace who marks feedback as helpful or unhelpful, so I propose there be no policy on doing such actions. --J (t) 16:39, 17 July 2012 (UTC)

You can see who marked it helpful and unhelpful. I disagree, I think having the policy encourages people to "do the right thing" and the community's view and consensus. Callanecc (talk • contribs) talkback (etc) template appreciated. 09:05, 18 July 2012 (UTC)

Oh. I didn't know that. Here's a proposed policy:

Marking comments as helpful or not helpful contribute to a feedback reviewer's decision. If you continuously mark the feedback as unhelpful, you will be told not to click the helpful or not helpful buttons again.

--J (t) 13:43, 18 July 2012 (UTC)

Sounds good what about this following the above:

Mark feedback as unhelpful if it does not contribute to the development of the article (such as "Justin Bieber is great"), but is not abuse and does not warrant hiding.

Mark feedback as helpful if it does not provide specific suggestions and/or only generally helpful (such as "more images please"). Marking feedback as helpful is the step below Featuring (for example "please add this specific image to the article")

What do we think of this? Callanecc (talk • contribs) talkback (etc) template appreciated. 14:17, 18 July 2012 (UTC)

If someone is following an edit filter or using Huggle to look at problematic users then they may only be seeing the unhelpful comments. Alternatively they might be skimming through a load of comments, ignoring anything that looks even vaguely plausible and just marking the ones that are definitely unhelpful as unhelpful and hiding the ones that merit hiding. What would be wrong with that? Recent changes and new page patrol both have patrollers who work that way, and if we make the mistake of implementing article feedback then we will need to recruit a load of volunteers to do this sort of thing instead of building an encyclopaedia. ϢereSpielChequers 07:30, 23 July 2012 (UTC)

Hiding/unhiding feedback

• Hiding is different from simply reverting, it is more akin to a revision deletion. Only Admins Rollbackers and Reviewers are able to see feedback once hidden. As such I think we should be more selective in hiding feedback then we might be in reverting talk page comments, especially considering that even gibberish comments still provide a good/bad rating. I would suggest hiding be used in the following cases:

1. Spam links
2. BLP Violations
3. Copyright violations
4. Personal attacks directed at specific editors
5. Grossly insulting, degrading, or offensive material, but not simple profanity not directed at anyone in particular
6. Personal information not rising to a level requiring full oversighting

My approach would leave a fair amount of chaff in the feedback list, but I think that is an acceptable result. The purpose should be to remove feedback that is actively harmful, not that is merely unhelpful. Monty₈₄₅ 21:35, 16 July 2012 (UTC)

Conditional support for the above, only if comments that do not make sense or are not relevant may be hidden. --J (t) 21:39, 16 July 2012 (UTC)

Frankly, I think that feedback consisting of "ehrkkkkkkk" or "hi" or "fuck you" should be hidden as well. David1217 ^{What I've done} 21:54, 16 July 2012 (UTC)

The way I look at it, a positive feedback that just says "hi", is still useful as the reader indicated they were happy with what they found. Likewise an unhappy result wit hthe string "ehrkkkkkkk" is also useful feedback as an unhappy result. Neither of those two textual responses cause any harm by leaving them in the system, other then taking up space on the page when you view it. As I advocate only hiding comments that are actively harmful, a line needs to be drawn as to when they start being harmful, personally I would draw that line such that mere profanity is not harmful enough to justify loosing the +/- feedback it is associated with, though I can understand someone may agree with my approach and still draw the line such that profanity is hidden. Monty₈₄₅ 22:07, 16 July 2012 (UTC)

I'm pretty sure that people who are responding "hi" and "ehrkkkkkkk" aren't happy or unhappy with the page, they're just adding random stuff. I've seen "hi" with an unhappy face and "ehrkkkkkkk" with a happy one. I'm fine with your approach though. David1217 ^{What I've done} 22:39, 16 July 2012 (UTC)

I'm going to go with David; why would you have random spam comments on pages? These comments are to help improve the page, not to express yourself however you want. --J (t) 23:14, 16 July 2012 (UTC)

My interpretation on the purpose of the feedback tool is that it is both for feedback suggesting improvement, and for feedback on the overall quality of articles. It is useful for us to know that some readers dislike the page even if they don't explain why or what we can do to improve it. Likewise, I think it is informative to see other comments not related to article improvement so we can understand if readers are getting what the article feedback tool is for. The comments are only visable to those reviewing the feedback special pages, and what they are interested in may vary greatly. Thats why I think the focus on our hiding the feedback should be on avoiding harm, rather then applying our view on what feedback is useful and what feedback is not. Monty₈₄₅ 15:14, 17 July 2012 (UTC)

On how to get this RfC out to people: should I add it to Template:Centralized discussion or one of the village pumps? David1217 ^{What I've done} 23:25, 16 July 2012 (UTC)

Yes. --J (t) 23:26, 16 July 2012 (UTC)

Back to topic: Should any nonsensical spam be hidden, regardless of AGF? --J (t) 16:29, 17 July 2012 (UTC)

Spam and nonsensical are totally separate. If it is actual spam it should be hidden, if it is just nonsensical it should be marked unhelpful, and left alone, as it may still prove useful to others, even if you don't have a use for it. Monty₈₄₅ 16:44, 17 July 2012 (UTC)

Should "OMG JUSTIN BIEBER I LUV U <3 LOLOLOL!!!!" be hidden? --J (t) 16:56, 17 July 2012 (UTC)

I really don't see the harm in leaving it. It tells you that someone appreciated the article about a subject they care about, and more importantly, it does no harm. When hiding a feedback from all other editors/readers, we should have a better reason then the feedback isn't particularly helpful. Monty₈₄₅ 17:03, 17 July 2012 (UTC)

WP:NOT#FORUM, and remember, this is the Article Feedback Tool, not the Article Forum Tool. --J (t) 17:18, 17 July 2012 (UTC)

I have no problem with the original text (at the top). I agree with Monty845 regarding not RevDeleting that sort of feedback as what David1217 & Jeffwang suggest is using the RevDel tool like rollback and it certainly is not. However I would encourage users (not just Monitors) to mark it as "unhelpful". Callanecc (talk • contribs)

---

I also propose changing the sentence: "Only Admins Rollbackers and Reviewers are able to see feedback once hidden." to "Only users called Monitors (admins, reviewers and rollbackers are able to view hidden feedback." Callanecc (talk • contribs) talkback (etc) template appreciated. 07:01, 18 July 2012 (UTC)

---

I suggest we also state when feedback should just be hidden and when it should be hidden as well as a user warning given. Callanecc (talk • contribs)

Requesting oversight of Feedback

• Feedback should be sent to oversight when it would qualify for oversight if found in other places. However, I would argue that contact information included in feedback can ordinarily be dealt with merely by hiding, if there is no indication they accidentally included it and there is no special circumstance that makes oversighting more critical. This is particularly true if the contact information is a free email address that likely reveals next to nothing about the person leaving the feedback if that is the only personal info they left. Obviously if there is a request from the individual for oversight of personal info, then it should be oversighted regardless. Monty₈₄₅ 21:43, 16 July 2012 (UTC)

  I'd like to get some existing oversighters, and also potentially WMF legal (the suppression policy is set by the board, iirc) to discuss here. I'll give them a poke when I land. Okeyes (WMF) (talk) 23:10, 16 July 2012 (UTC)

• The reason why emails and phone numbers should be suppressed is because anyone who gains the rollback or reviewer right can view it, yet they may not be identified by the WMF. Therefore, I have mapped out the following scenario:
• User A is an unidentified rollback, User B is a regular anonymous user, Rachel is the scammer, and User C is a debt collector.

1. User B posts his email address and phone number using the AFT5 tool, unknowingly its not a good idea.
2. User A hides it but puts it on a public database scam artists regularly use.
3. Rachel takes the information and calls this number.
4. User B thinks that Rachel is legitimate and Rachel gives User B fake debt services.
5. User C harasses User B
6. User B sends a cease and desist letter to User C.
7. User C refuses to stop and makes the harassment ever worse.
8. User B sues User C.
9. User C says that he got the info from Rachel.
10. Rachel says that she got it from User A.
11. User A is using one of WMF's websites, so User B sues the WMF.
12. The WMF is found guilty after a lengthy trial, even after pointers to the Communications Decency Act.
13. The WMF blocks User A and cannot sue User A, and so the WMF is in deep mud.

--J (t) 23:25, 16 July 2012 (UTC)

• Disagree - The tool does not ask too many questions and I feel is ideal for those people who do not want to interact too much with Wikipedia. Those who wan't feedback leave there comments in the talk page -Wikishagnik (talk) 03:17, 17 July 2012 (UTC)

No, comments are for improving Wikipedia, not expressing your opinion. Wikipedia is not a forum. Also, I can't understand what you're saying, are you supporting the oversighting of emails or are you against it? --J (t) 14:03, 17 July 2012 (UTC)

I am against any collection of personal information including emails. -Wikishagnik (talk) 12:22, 19 July 2012 (UTC)

• I think scenarios and hypotheticals about the CDA are probably best left to legal, which is not to say your example is wrong :). Can I suggest we put a hold on this until legal indicates that they're willing to allow for an exception? If they aren't, it's a moot conversation. Okeyes (WMF) (talk) 14:56, 17 July 2012 (UTC)

  Would it even be an exception? The relevant text of oversight policy is:

"Removal of non-public personal information, such as phone numbers, home addresses, workplaces or identities of pseudonymous or anonymous individuals who have not made their identity public. This includes hiding the IP data of editors who accidentally logged out and thus inadvertently revealed their own IP addresses. Suppression is a tool of first resort in removing this information."

Its not really non-public information if the person self discloses their contact information, and the policy is totally silent as to email addresses. We have adopted a practice of removing such information from requests at various desks/noticeboards, but afaik its not routinely oversighted, nor is there a policy that requires the removal of email addresses in particular, rather it is done as a courtesy. Monty₈₄₅ 15:06, 17 July 2012 (UTC)

Sure, but by that you could argue that any information posted on-wiki by the owner is immediately public and so should never be oversighted, and I'm fairly sure that's not the case in practise :). Having said that, my experience dealing with oversight as a volunteer is mostly limited to "12 year olds who don't get you shouldn't post your DOB on the intertubes", so I'll poke an OSer now. Okeyes (WMF) (talk) 15:28, 17 July 2012 (UTC)

Hi all, Okeyes asked me to comment here with my Oversighter hat on. Current oversight practice is (generally - everyone's mental guidelines are slightly different) that the line between "just redact" and "oversight this" lies between email addresses and phone numbers. Phone numbers can be traced back to people's real names, while most of the time, emails cannot (excepting email addresses like my_real_name(at)my_local_isp.com), and we will often oversight even self-posted phone numbers because people do not realize the privacy implications of posting them. Using this standard in the AfT will mirror usual community practice - emails are usually redacted/removed when they're posted onwiki (because people may not realize the spam implications), but they're not oversighted or rev-deleted most of the time, while phone numbers are quickly removed and oversighted. As far as I know, this is the rough standard that Oversighters have been applying to AfT oversight requests thus far. In most cases, whether someone has posted information pertaining to themselves or to someone else, we handle it the same way. (Note: I haven't been following this RfC, just popped in and gave it a quick read in response to Okeyes's request, so if I've missed the point people wanted addressed, please let me know what it was you were hoping to get answered and I'll try again) A fluffernutter is a sandwich! (talk) 15:42, 17 July 2012 (UTC)

Thank you Fluffernutter, that is very clear. So, to round it up, we should request oversight on emails that include names, addresses, and phone numbers? --J (t) 15:51, 17 July 2012 (UTC)

I read it as Addresses or Phone numbers go to oversight regardless, emails we hide, I don't think its clear about name + email, though I think we generally leave name without an associated email alone. Monty₈₄₅ 16:11, 17 July 2012 (UTC)

Here, we get a little bit into the realm of judgment calls, but I would say if someone posts an email address - or any feedback - that appears to contain their real name (when their account name is not the same or they're not open about their real name on WP) or other personal information (about them or someone else) like address, phone number, school they attend, etc, then please do request oversight on that feedback. We'd rather get a few too many requests, but be alerted to them so we can judge each on its own merits, than get too few and have potentially problematic information stay live. Also, you can always hide feedback and then send in an oversight request, to be safe - even if we decline the oversight, we have the ability to retain or remove the hiding part as we judge necessary. A fluffernutter is a sandwich! (talk) 16:21, 17 July 2012 (UTC)

Moving to new Level 2 header. --J (t) 16:23, 17 July 2012 (UTC)

Marking feedback as featured

Are the current guidelines OK? --J (t) 16:25, 17 July 2012 (UTC)

I think the best use of the featured option is to mark feedback that is a concrete enough that it will eventually be resolved one way or another. So feedback that identifies a specific issue, or that makes a specific suggestion for change/improvement. That way we wont end up with a bunch of featured feedbacks that we will never be able to mark resolved cluttering the system. Monty₈₄₅ 17:00, 17 July 2012 (UTC)

We need a guideline on this?

We need a guideline on how to respond to article feedback? Put another way, do we really need a guideline to govern everything an editor does on Wikipedia? I'm sorry, but this page is unneeded policy cruft. Ed ^{[talk] [majestic titan]} 09:04, 17 July 2012 (UTC)

Yes, we do need this guideline. Users can post some really bad things, and it is obvious that there is no uniform guideline (some users hide all caps even if it was made in good faith, which I revert), so yes, we do need a guideline. Right now, the oversighters don't want to suppress emails yet I see many AFT5 patrollers request oversight on emails, which I support them. --J (t) 14:00, 17 July 2012 (UTC)

Generally there are guidelines for the use of every specialized permission I can think of, this is no different (the ability to hide feedback is a specialized permission granted to admins, rollbackers and reviewers. Furthermore, the tool is generating a large number of oversight requests that are being denied, so it would be helpful for us to have some guidelines on what should and should not be submitted for oversight review. Monty₈₄₅ 14:36, 17 July 2012 (UTC)

Well, would that not better be incorporated into the Oversight policy? :) Okeyes (WMF) (talk) 15:10, 17 July 2012 (UTC)

It can be different from the regular oversight policy; it all depends on consensus. --J (t) 15:52, 17 July 2012 (UTC)

Well, the policies governing the OS tool come from the Meta policy actually; I don't know if ours has to follow that'n, which has to follow whatever the WMF say, or if ours follows just that'n, or if we can do our own thing, but I think it would be wise to find out, and I have given Philippe a poke. Okeyes (WMF) (talk) 16:06, 17 July 2012 (UTC)

Getting the public to know about the noticeboard

I'm proposing to add to MediaWiki:Articlefeedbackv5-header-message the following content:

"If you believe your comment was unjustly hidden and did not comply with [[Wikipedia:Article Feedback/Guidelines|guidelines]], please report the incident [[Wikipedia:Article Feedback/Noticeboard|here]]."

Just in case an IP didn't know how to get her comment back on the mainstream system, they can use the noticeboard to appeal. Anyone with the right to unhide comments may help process these cases; any registered user may clerk. --J (t) 16:18, 17 July 2012 (UTC)

Great idea, although how many people who have their feedback hidden have a problem with it. A great idea none the less. I suggest however that we do away with clerks for the noticeboard altogther, given that there are quite a large number of monitors I don't see the need to confuse the system with non-monitors. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:08, 18 July 2012 (UTC)

Oversighting issue

Moved from #Requesting oversight of Feedback.

Summing it up

• Emails: Oversight ones with names and personal info in their email addresses
• Phone numbers: Oversight immediately
• Any personal information: Oversight immediately if too revealing

--J (t) 16:27, 17 July 2012 (UTC)

General discussion

One point for clarification, most feedback is coming from IPs, not logged in users, so what should we do with real names from IPs? Monty₈₄₅ 16:32, 17 July 2012 (UTC)

We hide them and request oversight; IPs can be easily traced these days. --J (t) 16:35, 17 July 2012 (UTC)

Blocking Policy

It is far beyond the scope of the feedback guidelines to define new blocking rules. We don't block people as a punishment for making judgement calls we disagree with, particularly when there are steps short of block that can just as easily resolve the issue, such as removing a rollback/reviewer. The idea that you would sumarily block someone for a set period of time for hiding useful feedback 4 times is wholly inconstant with blocking policy and should be immediately removed from the draft. Monty₈₄₅ 16:39, 17 July 2012 (UTC)

Whatever. --J (t) 16:49, 17 July 2012 (UTC)

Absolutely agree. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:10, 18 July 2012 (UTC)

Removal of the ability to hide feedback

• Proposed language for removal of the ability to hide feedback:

The ability to hide feedback is inherited from other user permissions. If an editor is hiding feedback inappropriately, and the issue cannot be resolved, the ability to hide feedback may be removed using the normal process for removing the parent right that is granting the ability to hide. In the case of the reviewer or rollback right, any admin may remove the right, though it may still be prudent to start a discussion at a relevant noticeboard such as WP:AN/I.

I think this is the simplest way to deal with it, no need to forge new ground here. Monty₈₄₅ 16:40, 17 July 2012 (UTC)

Removing the AFT rights from rollback and reviewer is the best idea to this, otherwise rollbackers who were really helpful at reverting vandalism now loses a right he deserved to lose and a right he didn't deserve to lose. Also, there should be an AFT5 patroller usergroup. --J (t) 16:49, 17 July 2012 (UTC)

Generally an editor who is able to understand and follow the policies that govern rollback and reviewing should be able to understand and follow the policies developed to govern feedback. While I see no reason not to have a separate usergroup, I don't think such a usergroup is strictly necessary. Monty₈₄₅ 16:51, 17 July 2012 (UTC)

The only way to prevent someone from hiding useful feedback, while they have the ability to rollback or review PC pages, is to do the below. --J (t) 16:55, 17 July 2012 (UTC)

If they inappropriate use their user right (likewise with incorrecly using rollback) it is removed, there is absoluetly no reason to make any new policy. But perhaps a statement to that effect would be useful. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:12, 18 July 2012 (UTC)

AFT5Hide -> AFT5 Patroller

Proposed: Start an AFT5 Patroller requesting area at WP:PERMS, then rename the AFT5hide usergroup to AFT5 Patroller, next remove the AFT5 rights from rollback and reviewer, give the AFT5 Patroller to the people who commonly patrol AFT5 and finally give the right to all people who requested it and was successful in obtaining it at WP:PERMS. --J (t) 16:54, 17 July 2012 (UTC)

There's been an RfC that confirmed aft5hide as part of rollback/reviewer. Okeyes (WMF) (talk) 19:11, 17 July 2012 (UTC)

Oh nvm --J (t) 19:23, 17 July 2012 (UTC)

Hide none actionable comments?

Many of the comments are neither negative, resolved or actionable. For example "Thanks", "Nope" or "I love Justin Bieber". Would it be sensible to include those in the hiding feedback? Regards, SunCreator ^(talk) 23:42, 17 July 2012 (UTC)

We need to put ourselves into the IP's shoes. "Thanks" and "Nope" are okay, because it's the response to the given question, but "I love JB" needs to be hidden. --J (t) 23:44, 17 July 2012 (UTC)

Okay. Regards, SunCreator ^(talk) 23:47, 17 July 2012 (UTC)

I disagree with hiding any of those, what is the problem with "I love JB", mark it as unhelpful and move on. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:13, 18 July 2012 (UTC)

What is the problem? If you want to find good feedback, you have to sift through such irrelevancies.—Toccata quarta (talk) 07:43, 18 July 2012 (UTC)

P.S.: I love Nomen Nescio.

Hiding (have a look at the link) is a powerful tool and should only be used where necessary. There is nothing stoping you from marking it unhelpful, or as abuse. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:46, 18 July 2012 (UTC)

Unfortunately, I completely disagree with you, Callanecc. AFT5 is about feedback, and I love JB does not do anything for the article. Readers (and especially stupid Bieber flus) need to realize give us reasonable FEEDBACK or else we hide it. If you don't hide it, it will become a forum, which Wikipedia isn't. So, I've come to the conclusion to hide all feedback that isn't feedback. To Toccata quarta - that's called featuring. --J (t) 13:50, 18 July 2012 (UTC)

What about making use of the "mark as unhelpful" option? My position (and the position of the first users to discuss this - which is where I got my opinion) is that hiding should be be used sparingly and only when there is a clear vio of policy and and disruptive rather than not helpful for the article. Minor vandalism (your examples) shouldn't be hidden as, whilst they don't contribute anything ot the article, they don't do any harm (like advertising links do). While we have a less drastic and severe response tool (marking as unhelpful) we should use it; but we should still feel free to warn users (using the normal vandalism templates linking directly to their feedback). Callanecc (talk • contribs) talkback (etc) template appreciated. 14:27, 18 July 2012 (UTC)

The tool is about feedback. I see no reason for us to retain comments that don't relate to feedback. The entire purpose of this is to improve the articles, not to become a Justin Bieber fan club. Ryan Vesey Review me! 14:30, 18 July 2012 (UTC)

Well I take the point about Bieber fan club, but if we walk a line which each monitor can interpret their own way the line will become hazed. Unless there is a policy (etc) issue with the feedback it gets marked as unhelpful rather than hidden. But if the community would rather REVDEL everything which doesn't improve articles then we need to have very strict rules about what can and what cannot be hidden (ie what we class as a possible improvement and what we don't, otherwise we will be constantly dealing with monitors who cross the line. Callanecc (talk • contribs) talkback (etc) template appreciated. 16:14, 18 July 2012 (UTC)

As discussed in the above Hiding/unhiding feedback section, I strongly agree with Callanecc. In addition to sorting feedback by helpfulness rating to find the useful ones, directly actionable feedback should be featured to make sure its addressed. If we mark feedback that doesn't really provide useful information as unhelpful, the show helpful filter will allow you view just the helpful feedback. Monty₈₄₅ 16:24, 18 July 2012 (UTC)

Why don't we draw a line? Any non-useful comments like "hi" or "thanks" aren't hidden, but are rather marked as unhelpful, because they're neutral comments. Any comments like "OMG I LOVE U JB I WANNA KISS U !!!!!!" need to be hidden, especially if it becomes a discussion forum. --J (t) 16:46, 18 July 2012 (UTC)

Sample

Random sample, 171040 through 171049 (recent):

1. 171040 : (blank)
2. 171041 : i was lookinb ==
3. 171042 : mjihy758ew6tar
4. 171043 : i was looking detailed explanation in forward and inverse modeling in the interpretation of a potential field data
5. 171044 : (blank)
6. 171045 : by displaying some image. few existing bureau company
7. 171046 : NEVER
8. 171047 : more audio files
9. 171048 : nice actor
10. 171049 : (blank)

My opinion:

1. hide ok (revised 20:33, 18 July 2012 (UTC) because blank feedback is normally filtered out)
2. hide
3. hide
4. ok
5. hide ok (revised)
6. ok
7. leave it (probably trolling, but it is an answer to the question and we must assume good faith)
8. ok
9. hide
10. hide ok (revised)

benzband (talk) 19:47, 18 July 2012 (UTC)

I generally agree with this. In response to Jeff above, I still think comments like "hi" should be hidden; although, comments like thanks don't necessarily need to be. I dislike how we are treating this strictly as revdeletion because there is no alternative way to keep the material from being viewed. If someone writes "I LOVE JUSTIN BIEBER" in an article, I can undo the edit. Revdeletion isn't necessary because it is, for all intensive purposes, hidden. If there was an alternative, lets say collapsing a comment like "I LOVE JUSTIN BIEBER", I would be supportive, but as of right now I think hiding the material is the best option. Ryan Vesey Review me! 19:58, 18 July 2012 (UTC)

My version

1. Do nothing
2. Do nothing
3. Mark unhelpful
4. Mark Helpful
5. Do nothing
6. Mark Helpful
7. Do nothing
8. Mark Helpful
9. Mark Unhelpful
10. Do nothing

By hiding the six you suggest, we loose 4 feedbacks where the reader indicated they found what they were looking for, and 2 where they indicated they did not. Not the most critical information, but the text in the feedbacks does no harm, and so we are slightly better off keeping them. Nothing in there that justifies a quasi administrative action of hiding the feedback from public view. Monty₈₄₅ 20:03, 18 July 2012 (UTC)

Note we were using revdel as the standard because it was the closest community analogy. We can change it whenever there's a consensus-driven alternative :). Okeyes (WMF) (talk) 20:04, 18 July 2012 (UTC)

Personally, I wouldn't compare it to anything. The article feedback tool is so unlike any Wikipedia process that it should be treated separately. Ryan Vesey Review me! 20:05, 18 July 2012 (UTC)

But if someone wrote "I LOVE JUSTIN BIEBER" on their own user page or usertalk page, would anyone bother to blank it, let alone revdel or otherwise delete it? As you say, the feedback system is new, and doesn't really compare to anything else. Is it like an article talk page? Is it like a usertalk page? I think we should start out taking as few quasi administrative actions as we can till a clear problem is identified. Monty₈₄₅ 20:08, 18 July 2012 (UTC)

(edit conflict)Feedback is meant to improve Wikipedia. I doubt that editors who have come to the feedback dashboard to act upon these posts want to look through fancruft, abuse, spam or gibberish (i have encountered all of these). Many will probably just give up. benzband (talk) 20:24, 18 July 2012 (UTC)

• Lets consider another example Special:ArticleFeedbackv5/Ministry_of_Finance_(Turkey)/170482. Analyzing it, my first conclusion is that there is no major policy reason to hide or oversight it, it is also not actionable feedback for article improvement, and so should not be featured. Looking at the feedback we can speculate as to what the reader thought they were doing when they entered the text in the feedback box. They indicated they did not find what they were looking for (as indicated by the frowny face with the feedback). We probably are not going to add that information to the article, if that is what they are suggesting, but it is interesting to know someone went to the article looking for that info. Alternatively, we could speculate that they may have thought the feedback box was some sort of search function. I can imagine an editor looking to improve the experience at Wikipedia coming back through and seeing that feedback, and others like them, and using them for some as of yet unforeseen purpose. By hiding them, we foreclose the entire mental exercise of considering what the feedback means, as its now hidden from consideration. And for what? Monty₈₄₅ 20:19, 18 July 2012 (UTC)
  • I agree on that particular comment. I have taken a sample below of hidden-only comments, which you may wish to comment on. benzband (talk) 20:46, 18 July 2012 (UTC)
  • I actually disagree. In addition to keeping people from needing to wade through the comment, hiding the comments allows us too see when feedback is incorrectly used. I can look at all of the hidden content by choosing hidden from the dropdown filter menu at the Central Feedback Page. Then I can go through the comments and have much less searching to do to realize where to find it. I do wish there was a checkbox that could allow me to temporarily view all hidden comments rather than opening them each individually. Ryan Vesey Review me! 20:51, 18 July 2012 (UTC)

Hidden sample

Ten most recent hidden comments from last time i looked:

1. 171150 : He is Jewish!
2. 171143 : ebenezer yadoo 1500.fr èèuee gg
3. 171132 : biaecs hvjldgb jd;ljj c jhis
4. 171127 : gsjfkdzl.vkff\sya8'ahja. 6wwjdol 7ekdi7eqktywpfmnsdhrnfyrkrmrgeoe kdurtjfhweleithohkgtt;uljiltmruej.t
5. 171113 : hmmwsiucoheic vxuvhiosivkndyvlnu g hit nvihdkhf8nihnh hercohijighiv hivhjvuhv mvlvorv8nhnihkhjunivbipvhuvjyfksjmnlkbcbjcuunhb hhvjndbih8 vuvjun gyhwigi focihpojbhhorhfiuuhfuhhiourilwkovujlkumkvfkljgkvmkmluhcihcuugugcygugdygcbh7cucyu88hefohkhihjfihf hiuhihpocjojfjkkoiuiheiuhwqjtnkj m ndjvdd jchuhjncqkqjjqjqjjqhhqwksk,f7kchldilhFOHXTGHPWHGIPnckdihichgokjjiiuiiiuiiiiiiiuiiiuiiiuiiiuloljgjjo9hjbojojivrji h bjivnjhmgjivimgik mmis odama is the pnesidont of u.s.a. . jhujwovphvhucojqjjakzkfp;dlkhvn/hdyhiugd.fhn irugvjufjuhfhbohuiijckiubgohedowowjvnii;wdgjgfinihniodjnviihoghjnviiunbpinoihgpihhba;Hhbbibkjhighihkg;goivnuhfjefhgyhjvhfudjjxkkj wwqppfie jdnjhdhhggfjjyjrhhjhivbaixnuhckdjfhkjdehfiehedig
6. 171105 : kjbb mjy nnvun ;lkjj;;pklknkfgklj;l[pl;;oojpik
7. 171098 : U are a ass hole and bitches
8. 171087 : Fuck You
9. 171071 : Dear Sir, Dear Madam, We are interested in leasing MQ-9 Repear, if not possible, in importing Honeywell TPE331-10 turboprop, 900 hp (671 kW) with Digital Electronic Engine Control (DEEC) for Turkey. Please let me know we could develop business with your esteemed selves. Thanks in advance <Name, Business Name, and Street Address redacted> Fort Lauderdale, FL 33311
10. 171060 : shit

— benzband (talk) 20:42, 18 July 2012 (UTC)

I would have hidden nine of those. "He is Jewish!" depends on the context: if the subject has been described as Jewish in secular sources then it is relevant feedback, whereas if the intention appeared anti-semetic then it was rightly hidden. —WFC— 00:50, 19 July 2012 (UTC)

No consensus?

What we can agree on:

• RevDel isn't like AFT5's Hiding.
• Anything unhelpful should be marked as so.

What we can't agree on:

• Neutral comments that don't help the article or act like spam should be (deleted/not deleted)
• Which to hide?!

--J (t) 03:47, 19 July 2012 (UTC)

• I would clarify the first point and say that AFT5 hiding isn't revdel, but it also isn't regular undoing. Second and third points I agree. I would add also that I think we all agree that certain things, including oversightable material, BLP violations etc.. should definitely be hidden. Monty₈₄₅ 04:00, 19 July 2012 (UTC)

• I think the third and fourth points amount to the same thing. The disagreement is surprisingly narrow: posts which any reader would consider to not be helpful and intended not to be, but which on the other hand probably wouldn't be reverted if posted on a talk page. Ordinarily I would simply say that this should be a matter of discretion, but if we are talking about blocks and removing rights, it is important that the line is as clear as we can make it. —WFC— 04:19, 19 July 2012 (UTC)

Perhaps we could get an RFC going?—Toccata quarta (talk) 05:58, 19 July 2012 (UTC)

There is already a general RFC open at the top of the page, though I have no objection if someone wants to open a second, narrow RFC on the specific question. Monty₈₄₅ 13:03, 19 July 2012 (UTC)

Featuring Feedback

Should be add something about when to Feature feedback? How does the following sound:

Feedback should be featured it contains specific tips and detailed ways to improve an article - for example "Please add to the article Angry Birds that the app uses Lua to save its settings." If the feedback meets this requirement and does not contain anything which meets the conditions for hiding or otherwise might be considered offensive, please feature it.

Comments and discussion welcome, Callanecc (talk • contribs) talkback (etc) template appreciated. 07:58, 18 July 2012 (UTC)

I agree. And thanks for using my example. Did you source it correctly, because we can't have copyright violations! --J (t) 13:45, 18 July 2012 (UTC)

Actually I was thinking of changing it to something a little less specific (what do you think): "Change the 'Aims of the Organisation' section from a list into prose." Callanecc (talk • contribs) talkback (etc) template appreciated. 14:20, 18 July 2012 (UTC)

My view is that feedback should be featured if it is reasonably actionable, and at some point is likely to be marked resolved either as done, or as a rejected suggestion. It shouldn't need to be entirely specific, just specific enough that there can be a definite response. So I would say feature "This article could use more pictures", as it can be resolved by either adding more pictures, or after review of the article, by deciding it has enough pictures. Conversely, something like "This article is great" or "This article sucks" isn't actionable, and so featuring it for further review wont really result in an outcome. Monty₈₄₅ 20:25, 18 July 2012 (UTC)

Caveats

How does something like the following, as it's own section sound?

• In general, an editor should not use any auto/confirmed (for example, featuring feedback) or monitor (for example, unhiding feedback) tools on feedback they submit.
• Once an action is undone it should not be redone, except where doing so would unquestionably meet one or more of the conditions above.
• Once a feedback item has it's initial Oversight request declined, Oversight should not be re-requested. If a monitor feels that it should have been Oversighted they may request this at Request for Oversight, or by re-requesting Oversight using the tool with a detailed reason.

What do we think of the above? Please feel free to comment (especially on the Oversight caveat). Callanecc (talk • contribs) talkback (etc) template appreciated. 08:25, 18 July 2012 (UTC)

Could we not just apply the principles of WP:EW or WP:WHEEL to this feature? ItsZippy ^{(talk • contributions)} 19:10, 19 July 2012 (UTC)

---

Remove the oversight requested and oversight denied filters?

Moved to Wikipedia talk:Article Feedback Tool/Version 5 § Remove the oversight requested and oversight denied filters?

After playing with the feedback tool for 3 minutes I realized I can filter for oversight requested and oversight denied. These comments are hidden, but still available for any rollbacker to view. This is completely different from the way oversight is handled through the rest of the encyclopedia. Please remove these filters or make them avaiable to oversighters only. Yoenit (talk) 10:50, 18 July 2012 (UTC)

• Disagree - firstly, it is really easy to accidently press the oversight button and then you have to unrequest it. This post was inappropriatly requested for oversight. Not to mention the tool can be misused. And why must you hide "oversight declined"? They are declined, right? CyanGardevoir ^{(used EDIT!)} 11:20, 21 July 2012 (UTC)

• As the move template states, the conversation has been moved to Wikipedia talk:Article Feedback Tool/Version 5#Remove the oversight requested and oversight denied filters? so please post your comment there. Thanks, Callanecc (talk • contribs) talkback (etc) template appreciated. 11:28, 21 July 2012 (UTC)

Foreign language responses

There is a fair proportion of foreign language feedback. What I have seen does look about as relevant and helpful as the English feedback, Therefore I recommend that we remove the hiding non English part of the proposed guideline. Just let the other criteria apply. Graeme Bartlett (talk) 11:01, 19 July 2012 (UTC)

Sounds good to me, I've been using Google Translate to check whether the foreign language feedback is ok or not. Callanecc (talk • contribs) talkback (etc) template appreciated. 11:17, 19 July 2012 (UTC)

I was thinking the same. Not all foreign feedback is unhelpful and should not be hidden just for not being in English. Is it possible for a 'non-English' flag to be created, so that monitors can flag feedback as non-English and people who know those languages can help translate them? ItsZippy ^{(talk • contributions)} 13:18, 19 July 2012 (UTC)

Interesting idea but where would we put it? Can we do feedback on feedback, or have a talk page for the feedback, or even add to the same feedback? I don't think that is there yet, but some of those buttons have a comment field that could be used for a translation, perhaps feature, as if someone goes to the trouble to translate or understand, and it is something worthwhile then featuring with translation is the way to go, else click as unhelpful, abuse or hide if that is suitable. Graeme Bartlett (talk) 21:27, 19 July 2012 (UTC)

I have no in-principle objection (although a did hide a four-letter post in non-latin script earlier, based on the guideline and the unlikelihood that four letters could convey anything constructive). Foreign-language wikis accomodating English speakers to a limited degree has helped me out in the past, and within reason I think this is an area where we can and should return the complement. My one concern with not filtering foreign language content is if we are unable to determine its meaning. Tools such as Google translate can help detect profanity in most languages with a significant speaker count, but it might prove a lot harder to tell between a constructive but critical comment and a BLP violation. —WFC— 17:54, 20 July 2012 (UTC)

I think in this case it needs to be an 'if Google Translate' (etc) can't give a definite response, we will need to lean towards the worst case and mark as abuse if not hide. Callanecc (talk • contribs) talkback (etc) template appreciated. 18:14, 20 July 2012 (UTC)

You really can't mark something as abuse simply because you don't understand the language. That would be a recipe for newbie biting. ϢereSpielChequers 07:41, 23 July 2012 (UTC)

Addition on 20 July

Based on the discussion above I created a draft version of the Guidelines, and the following to allow for some discussion generally, on individual section and further sections which need to be added. Feel free to add further level 3 & 4 headings in this section as required. Callanecc (talk • contribs) talkback (etc) template appreciated. 17:26, 20 July 2012 (UTC)

General discussion

Lead section

Featuring feedback

What is the general opinion regarding adding another reason to feature so that the feedback to be marked as resolved. Such as if what is requested is already in the article, or if it is a reasonably suggestion, but there is absolutely no chance of it happening in the article it's suggested for. Callanecc (talk • contribs) talkback (etc) template appreciated. 18:39, 20 July 2012 (UTC)

I'm in favor of advancing as many comments as we can to the resolved stage. If the request is already done, you do it immediately, or its a reasonable suggestion but you reject it for a policy reason, I think its fine to mark it featured and then immediately mark it resolved with the explanation of why its resolved. Monty₈₄₅ 18:53, 20 July 2012 (UTC)

 Done Callanecc (talk • contribs) talkback (etc) template appreciated. 12:32, 21 July 2012 (UTC)

Marking feedback as un/helpful

Hiding and unhiding feedback

I would like to remove "Grossly" and add "this criterion should be interpreted liberally" to WP:AFT5#H2. Given that we don't have the option to revert comments, I think we shouldn't be as strict with hiding comments as you would using Revdel. What is the community's opinion on this? Callanecc (talk • contribs) talkback (etc) template appreciated. 12:31, 21 July 2012 (UTC)

 Done If anyone doesn't agree feel free to undo and discuss here. Callanecc (talk • contribs) talkback (etc) template appreciated. 03:38, 22 July 2012 (UTC)

---

I very strongly think that feedback like "asdkjfha;sdkfjafd" and "fuck you bitches poop poop poop" should be hidden. They are bad-faith comments that are intended to cause disruption. Monty845, you say that we need to keep the feedback visible to see whether they were happy or not with the page: I disagree. If a user is looking through feedback for ways to improve a page, saying "I'm happy" (which is pretty much what "asdkjfha;sdkfjafd" with a smiley face is doing), is next to useless. The disruptiveness of the comments outweighs the very little helpfulness they provide. David1217 ^{What I've done} 16:43, 21 July 2012 (UTC)

I agree, but the key difference between uninformative positive and malicious feedback is that the latter deserves to be replied to with a warning/block, whereas the first does not.—Toccata quarta (talk) 18:44, 21 July 2012 (UTC)

I don't like feedback such as ":-) great!" and ":-( too short". However, to filter comments like this for any reason other than a spam-like volume from one person is to step decisively into the realm of censorship. As a side note, even if they are submitting a lot of them, it's worth checking to see whether they are accurately identifying good and bad articles. If so, it may just be that they are trying to help us assess articles. Explain to ten people in this situation how to do this, and a couple might become users. —WFC— 18:54, 21 July 2012 (UTC)

I think the line between unhelpful and hidden is "I <3 JUSTIN BIEBER!!!" and "fuck you". The former, while unhelpful, is not bad-faith, while the latter is. David1217 ^{What I've done} 21:05, 21 July 2012 (UTC)

David, I agree this comment. But I would say that "asdkjfha;sdkfjafd" should be marked as unhelpful and not hidden, but "fuck you bitches poop poop poop" should be hidden and would already meet both H6 & H2 (just). Callanecc (talk • contribs) talkback (etc) template appreciated. 03:38, 22 July 2012 (UTC)

In H6, I would like to change 'This includes allegations..." to "This is limited to allegations...". To make it clear that things like "asdkjfha;sdkfjafd" should be marked as unhelpful (as stated in "Marking feedback as un/helpful") but not hidden. Callanecc (talk • contribs) talkback (etc) template appreciated. 03:43, 22 July 2012 (UTC)

Without wishing to imply that I intend to deliberately ignore this guideline (I will abide by my understanding of community consensus), I suspect that in general editors will hide unambiguous spam whether we tinker with the verb or not. All this change would do is create drama at a later date, which is precisely what trolls would want. —WFC— 04:08, 22 July 2012 (UTC)

IMHO, gibberish feedback should be hidden (but of course i won't do it if without consensus). benzband (talk) 08:01, 22 July 2012 (UTC)

While I'm in the not hiding camp, I don't see much consensus one way or the other, if anything it may lean slightly in favor of hiding. Need more participants to weigh in on the issue. Monty₈₄₅ 14:28, 22 July 2012 (UTC)

I agree. I don't want to dominate this discussion, but will add one further thought. My understanding is that approximately 11 in 12 articles don't have the tool enabled yet, and in my opinion the post-rollout volume of "asdkjfha;sdkfjafd" type posts would be a relevant factor in deciding which side of the fence to come down on. —WFC— 20:12, 22 July 2012 (UTC)

Based on how much "asdkjfha;sdkfjafd" feedback we're getting now, there'll be a hell of a lot once most pages have it. David1217 ^{What I've done} 23:57, 22 July 2012 (UTC)

David & WTC you make good points, and it has moved me from 'against' to 'on the fence', the main reason I don't support it is that I am hesitant to treat hiding like rollback, but am open to be convinced. Callanecc (talk • contribs) talkback (etc) template appreciated. 02:07, 23 July 2012 (UTC)

Yep I've changed my mind and since I was the only one really opposed I've edited the guidelines to make it clear that this type of feedback may be hidden. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:53, 25 July 2012 (UTC)

 Done Callanecc (talk • contribs) talkback (etc) template appreciated. 07:59, 25 July 2012 (UTC)

I've moved personal attacks from H3 to H2, as that is the sort of thing H2 is talking about. If anyone doesn't agree feel free to revert and/or discuss. Callanecc (talk • contribs • logs) 12:02, 7 August 2012 (UTC)

Requesting oversight

Conflict of interest

Wheel warring

Can someone explain the need for inconsistency in this paragraph?

Sanctions will vary depending on whether the users involved are admins or reviewers/rollbackers. Admins are subject to the conditions at WP:WHEEL, so usually in an immediate Request for Arbitration. Reviewers and rollbackers may have these rights removed by an uninvolved admin; whilst a discussion at Administrators' noticeboard/Incidents is not required, it is encouraged.

I think pre-emptive removal is inappropriate for less clear-cut cases (I'm primarily thinking about reinstating an action once). An admin would absolutely never be desysopped for a single reinstated admin action before the conclusion of an Arbcom case. On the other hand an admin might be temporarily desysopped by a crat if they were engaged in continuous, current wheel-warring. I would therefore suggest changing the paragraph to:

Sanctions will vary depending on whether the users involved are admins or reviewers/rollbackers. Admins are subject to the conditions at WP:WHEEL, so usually in an immediate Request for Arbitration. Reviewers and rollbackers may have these rights removed by an uninvolved admin – whether removal comes before or after discussion at Wikipedia:Administrators' noticeboard/Incidents would depend on the extent of the wheel-warring.

The new wording gives admins enough latitude to remove rights where they consider temporary removal to be necessary, while at the same time being clear enough that removal should not be done lightly. —WFC— 18:28, 20 July 2012 (UTC)

 Sounds good to me, since I came up with that wording on the spot, I've changed it yours. Callanecc (talk • contribs) talkback (etc) template appreciated. 18:33, 20 July 2012 (UTC)

Further sections

Name of the page

I think we may want to rename the page, there is a separate page at Wikipedia:Feedback guidelines which provides instructions for leaving feedback. To avoid the confusion of both that page and this one being essentially the "feedback guidelines", we should come up with a more descriptive/distinct name for this page. I'm not exactly sure what that should be though. Monty₈₄₅ 00:59, 23 July 2012 (UTC)

Feedback response guidelines? Callanecc (talk • contribs) talkback (etc) template appreciated. 02:00, 23 July 2012 (UTC)

 Already done by User:Monty845. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:58, 25 July 2012 (UTC)

Demote inanities to likes/dislikes

I'm not a fan of AFT as I see it as at best undermining SoFixIt and at worst diverting editors from useful stuff to trawling through an endless stream of +1s and "is gorgeous" comments. I doubt that AFT can be made a net positive, but there is a way in which the timesink nature of AFT could be reduced. Create subcats of Article feedback for likes and dislikes and allow the patrollers to classify inane comments accordingly. That way only one AFT patroller need ever look at a particular plus 1, Thanks Wikipedia! Or Luv ya JB:) but we won't be biting the readers who've made those comments. Remember the original idea behind this was to engage the readers, if we then handle the consequent tidal wave of inanity in a bitey abrupt way we render the whole thing counter-productive. Hopefully some kind bot writer will then be able to automate the processing of common like and dislike comments so that no editor need actually look at them. ϢereSpielChequers 07:59, 23 July 2012 (UTC)

Do you mean "distinguish by how helpful the comment has been rated" or "distinguish by whether they are reporting a positive or negative experience"? And yes, we've heard your SoFixIt theory, and we've tested it, and we've found that AFT doesn't cannibalise...so that's not really an argument that can be made at this stage :). Okeyes (WMF) (talk) 11:05, 23 July 2012 (UTC)

Glad to hear you've tested it. Please can I have a dif to that test. ϢereSpielChequers 14:34, 23 July 2012 (UTC)

Sure! See m:Research:Article feedback/Stage 3/Conversion and newcomer quality. Either no cannibalisation happens or (more likely) there is a bit of cannibalisation, but it's drowned out by the much larger number of people who find out they can edit and then try to. Okeyes (WMF) (talk) 14:59, 23 July 2012 (UTC)

An interesting start, and I'm not surprised that a call to action did so much better than AFT at getting more editors. As compared to the control, I'm not convinced that the test has sufficiently allowed for the sheer novelty value of either AFT or the much more successful call to action. But I'll make my comments on the appropriate talkpage for the test. ϢereSpielChequers 15:39, 24 July 2012 (UTC)

These tests were conducted about four months after we first turned AFT5 on. I think novelty has probably worn out at that point for a lot of people. Okeyes (WMF) (talk) 15:54, 24 July 2012 (UTC)

Merging H2, H3, and H6 (and one other suggestion)

H2 and H3 are quite similar; a "personal attack" (H3) is certainly "insulting, degrading, or offensive material" (H2). The same thing applies with H6—"insulting, degrading, or offensive material" and "personal attacks" are certainly "purely disruptive material that is of little or no relevance or merit to the project" (H6). However, I propose that H3 is merged into H2 (and renumbered H3 for consistency with WP:RVDL) and the note kept that "this criterion should be interpreted liberally". Also, I propose changing H2 be feedback that isn't as disruptive as personal attacks or insulting, degrading, or offensive material (i.e. gibberish or spam). Any thoughts? David1217 ^{What I've done} 18:17, 25 July 2012 (UTC)

There's been no response after two days, so I'm being WP:BOLD and carrying out my proposed changes. David1217 ^{What I've done} 16:23, 27 July 2012 (UTC)

 Already done Looks good to me. Callanecc (talk • contribs) talkback (etc) template appreciated. 11:56, 28 July 2012 (UTC)

Constructive feedback that also contains an email address

While reviewing flagged feedback to hide those containing email addresses, the question has come up of what should be done about a constructive feedback that also contains an email address. So far I've copied a redacted version of the feedback to the relevant article talk page, explaining its source and crediting the IP, while hiding the email containing feedback. I figured it would be a good idea to raise it here to see if its agreed that doing so is acceptable, and assuming it is, to consider whether we should provide guidance to encourage doing so.

On a somewhat related point, there is also the people attempting to contact an organization, and leaving an email for a response. Should we be doing anything, such as leaving the IP a template response (that they may never read), explaining how we aren't able to help them? It just seems like a bad idea to have the feedback end up in black whole when the poster seems to be expecting a response. Monty₈₄₅ 00:27, 26 July 2012 (UTC)

This is an important point. This is one example of it, and I'm not quite sure what to do with it. Toccata quarta (talk) 19:48, 1 August 2012 (UTC)

I'm going to have a go at coming up with a single issue notice for personal info and this over the next few weeks. Callanecc (talk • contribs) talkback (etc) template appreciated. 07:32, 4 August 2012 (UTC)

The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.

User warnings

Hi all, the user warnings (levels 1-4 & 4im, adding personal info, as well as trying to contact the article subject with the feedback) are done and added proper template syntax, if everyone would mind having a look. Callanecc (talk • contribs • logs) 08:04, 13 August 2012 (UTC)

For the record I won't be using those often. Take away the glory of attention, hide every trace of the original activity (or at least, make sure that the vandal can't see it), and there's very little motive to carry on vandalising. —WFC— 11:36, 13 August 2012 (UTC)

Same here. benzband (talk) 11:54, 13 August 2012 (UTC)

Hmm. While I thing that the warning templates are good (thanks for the work, Callanecc!), I don't see many repeat offenders, and so warning them will probably actually have a negative effect ("Ooh, I got a reaction! Let's try again!"). Perhaps instead, we should change the line under the feedback box from "Please post helpful feedback. By posting, you agree to transparency under these terms." to something like more stern (although WP:BEANS?). David1217 ^{What I've done} 19:10, 14 August 2012 (UTC)

I agree that there aren't many repeat offenders, but for the ones that are (and there are some) it would be helpful to have something we can do about it. I imagine that the two single issue notices will get a work out though. Callanecc (talk • contribs • logs) 02:23, 15 August 2012 (UTC)

That's a good point, but I would go as far as recommending against creating levels one to four, as I am certain that they will create a problem which doesn't currently exist. When feedback is hidden there is no evidence, and this alone normally deters repeat offenders. Adding a series of polite, easily gamed warnings would simply encourage IP hoppers to persevere.

Assuming that we are still working from here, the 4im warning looks fine. The personal information one is well worded, but should only be used if the information has been oversighted, given that non-identified users can see ordinarily hidden posts. The contacting one seems a little bit abruptly worded though. In my experience those who ask for contact details generally do so because they do not fully understand the purpose of a Wikipedia biography, and most of the time are asking in good faith. —WFC— 20:54, 17 August 2012 (UTC)

Sounds good regarding getting rid of levels 1-4. Regarding persinfo oversight or if it's hidden. I'm planning to re-write the contact one to make it sound a bit 'nicer' and to include a message about personal info (as that's when people generally include personal info. Would you mind leaving a comment at User talk:Callanecc/sandbox/AFT5 outlining your comment? Thanks, Callanecc (talk • contribs • logs) 13:38, 18 August 2012 (UTC)

Marking as un/helpful ... generic compliments?

Can someone please add instructions about whether generic compliments like "nice informative article" should be marked as helpful or unhelpful? Obviously they don't help us much to specifically improve the article, but they're genuine feedback and will encourage editors. --99of9 (talk) 10:52, 15 August 2012 (UTC)

So I guess we should mark them as helpful but then also mark them as resolved?? --99of9 (talk) 11:35, 15 August 2012 (UTC)

I'd just ignore, it is isn't really helpful and it hasn't been resolved. Callanecc (talk • contribs • logs) 11:45, 15 August 2012 (UTC)

Ignoring it is likely to produce pages and pages of stuff you'll have to sift through every time you want to check for unaddressed feedback. I'd say "resolve" it (put "thanks!" or some such in the comment line) and get it off the radar screen! MeegsC (talk) 18:41, 17 August 2012 (UTC)

Good point, what is everyone else's opinion on this? Callanecc (talk • contribs • logs) 13:42, 18 August 2012 (UTC)

I suspect that the WMF would have planned on using data relating to resolved comments, and for us to mark irrelevant stuff as resolved might hinder them. WMF always leave decisions like this to the community, but if we were to inform them that as things stand we are likely to do what MeegsC proposes above, but that we would prefer to put these comments in a separate category, they might be willing to create a technical solution. —WFC— 16:35, 19 August 2012 (UTC)

A "Thanks" option which also marks as resolved? (Such compliments are by technicality resolved.) WMF can still mine the data appropriately then. --Izno (talk) 01:51, 21 August 2012 (UTC)

What about unhelpful generic criticisms? Example: "[Subject of article] is dumb." I'd prefer to just hide comments like that, but I'm having a hard time justifying that because it's not really an attack against anybody in particular. It's just an unhelpful opinion. ~Adjwilley (talk) 01:59, 21 August 2012 (UTC)

I think H2 covers that, no problem (the "apply this liberally" at the end seems to cover it). It's what I've been doing. --Izno (talk) 02:10, 21 August 2012 (UTC)

Ok, Thank you. ~Adjwilley (talk)`

Yep, that's what H2 is for, however you can also use H5 (if the subject is a living person), or H3 ("allegations, harassment"). Callanecc (talk • contribs • logs) 08:43, 21 August 2012 (UTC)

Sorry to revive older discussion but has been anything new on this matter? That i see fair amount of feedback on the lines of positive feedback: "As good as it gets for my needs, very informative, thanks." or positive feedback: "No improvement great". Guidelines for these are not exactly well defined and those are - on brief glance - the most common type of feedback. I thought of tagging them as 'unhelpful' and 'resolved' since neither provide any insights into how the article in question should be improved nor do they necessitate any changes - but after reading this page i think that would wrong approach to the issue. - Wanderer602 (talk) 20:36, 25 September 2012 (UTC)

Thanks for the message on my talk page, I've added a bit to the section on marking as resolved (which I believe was where the consensus went) to make it clearer what to do. Callanecc (talk • contribs • logs) 05:04, 7 October 2012 (UTC)

Suggestion re. e-mail

I just reviewed a couple pagefuls of recent feedback for the first time, and noticed that a substantial fraction of the meaningful posts (excluding patent nonsense, which seems to comprise the majority) include e-mail addresses. Since feedback is solicited from visitors who are not expected to be aware this is considered problematic here, I think it is inevitable that this will happen a lot, even in feedback that is actually constructive. My suggestion:

Would it be possible to programmatically scrub e-mail addresses (anything syntactically resembling one) from the text when posts are submitted, automagically, without the intervention of human moderators?

If the principal objective of the AF tool is to tease readers into becoming contributors, flagging their post as abuse and/or blanking the whole post is not very enticing. ~ Ningauble (talk) 22:44, 18 August 2012 (UTC)

That would be great! I agree with Ningauble that much of the feedback is helpful, but has an email address in it. Okeyes, is this feature possible? David1217 ^{What I've done} 22:57, 18 August 2012 (UTC)

The best solution in my opinion is to edit the feedback (to remove the e-mail address from there) and give a reply on the contributor's talk page. That would prevent the dissemination of spam. Toccata quarta (talk) 04:24, 19 August 2012 (UTC)

I agree with Toccata quarta, being able to edit feedback would be the best solution. Someone may want to either move this discussion to the main AFT talk page, or at least drop a link to it there, as its more of a feature discussion then a policy one, and it will probably get more attention there. Monty₈₄₅ 04:36, 19 August 2012 (UTC)

I brought it up here because I noticed the RfC at WP:CENT, and because a significant amount of feedback response activity seems to involve H4 – exposure of e-mail addresses. I apologize if it was off-topic to bring it up here. Chalk it up to a muddleheaded failure to recognize the distinction between responding to a problem (the topic here) and trying to obviate it.

I notice there is already a related thread at Wikipedia talk:Article Feedback Tool/Version 5#e-mail address, where a WMF representative recommends against built-in e-mail suppression and suggests using the Abuse Filter (which would reject the entire post).

Regarding David1217's observation that I said much of the feedback is helpful: that is not what I meant to say, and I would not say so. Rather, I would say that in my sampling of feedback the signal to noise ratio is very, very low. I apologize for being so inarticulate as to give the wrong impression about what fraction of the feedback appears helpful to me.

The only reason I commented at all is because the fraction of feedback that is actually constructive and potentially actionable already appears to be miniscule before throwing away the ones that inappropriately include contact information (and branding them with a scarlet A for abuse). That is just my feedback on the process of responding to feedback. ~ Ningauble (talk) 15:43, 19 August 2012 (UTC)

Just to clear up something, it isn't possible to edit (and isn't likely to happen, is it would require a major rewrite of the software) feedback, and in this case it wouldn't work, because the email address (etc) would still be available in the submission history (as you can't delete the first version - which has an email address in it). In terms of the software automatically hiding/deleting email addresses. If the feeback containing the email address is useful then (I believe) it has already been suggested (either here or WT:AFT5 - can't remember which) that the reviewer put the feedback on the article's talk page. Callanecc (talk • contribs • logs) 11:50, 20 August 2012 (UTC)

I appreciate that these things are never as simple as they sound, but I doubt adding a step to preprocess the text before it is posted would greatly disrupt the architecture. ~ Ningauble (talk) 20:57, 20 August 2012 (UTC)

• Ultimately I think this all comes down to "do people think the risks of allowing email addresses outweigh the benefits?". If yes, then we should talk about using the abuse filter. If not, we shouldn't. Okeyes (WMF) (talk) 14:34, 20 August 2012 (UTC)

  I think the risks are various:

  1. People providing or requesting information, in a constructive and useful way, who naively expose personal contact information in a very public place. It might be argued that we have no business protecting them from themselves, but it should be noted that the general readership, including children, may often have no idea what a hotbed of troublemakers the inside of Wikipedia can be.

     By way of comparison, there is nothing preventing someone from openly posting contact information on their user page or in a discussion but, unlike the feedback system, if they later decide it was a mistake they can re-edit the page.

  2. Spammers and phishers using this as a venue for baiting dupes.
  3. Impersonation and other malicious activities that come to mind but aren't worth mentioning.

  The advantage is fairly straightforward:

  1. Allowing people who wish to provide or request information to initiate off-wiki communications for that purpose. It may be debatable how constructive this is as a step toward the AFT objective of enticing readers to become contributors.

  If the benefit outweighs the risk then proposed guideline H4 should be deleted and moderators should stop flagging them as abuse. If the risks outweigh the benefit then the Abuse Filter is not the only option, and it may be worth considering options that do not involve throwing out the whole post (baby) with the e-mail (bathwater).

  Please note that Wikipedia:Feedback guidelines#What to avoid already says not to include e-mail addresses, and that the software will reject them. However, this does not appear to be happening, at least if it isn't a live "mailto:" link. ~ Ningauble (talk) 20:57, 20 August 2012 (UTC)

  IMHO, the risks outweigh the benefits. benzband (talk) 21:16, 20 August 2012 (UTC)

  It seems to me that an input could be added to the form which allows for email, and only having certain classes of user be able to view it (just as e.g. monitors can only see the "hide" button). If only for OTRS members (I would argue that we can also trust those in the monitor group [rollback, admin, reviewer]), having a dedicated form would allow trusted users to answer concerns with appropriate followup, which in this implementation, I think would negate the negatives and (at the least) go nowhere on the singular positive, if not advance the positive. I'm not sure this is possible or too incremental of an update, naturally. --Izno (talk) 17:00, 21 August 2012 (UTC)

  Alternatively, and more simply, we could provide links to other resources such as Help desk, Teahouse, or OTRS. However, that might clutter the input form with too many distractions. ~ Ningauble (talk) 17:53, 21 August 2012 (UTC)

• If you really want to support email responses, it should be done through a feature like "email this user", i.e. the person enters their email address into a form field and it's saved in the database but not exposed to editors. The software would have to send the person a confirmation url to make sure the address was really theirs. 69.228.170.132 (talk) 17:54, 22 August 2012 (UTC)
  • Why do we want email addresses at all. Most, if not all, of the feedback which includes an email address is a request for the subject of the article to get back in contact with them (which is just not going to happen). Do we really want to encourage people to leave their email address at all? Callanecc (talk • contribs • logs) 04:51, 23 August 2012 (UTC)

    Telling people not to post e-mal addresses is longstanding procedure at WP:HELPDESK, is also done at the newer WP:Teahouse, and has been in Wikipedia:Feedback guidelines ever since the first version posted by Okeyes (WMF). I think the answer to Okeyes' rhetorical question above is pretty evident: people don't think e-mal addresses are a good idea even in contexts that are more engaged than this one. ~ Ningauble (talk) 16:11, 23 August 2012 (UTC)

    And that is done because that information is then completely public and released to the world and Internet at large, and not the assumed reasoning of "good idea in contexts of higher engagement". That does not need to be the case with AFT, as I (at least) have suggested. --Izno (talk) 00:34, 24 August 2012 (UTC)

Marking feedback as "RESOLVED" very confusing

It seems to me that is highly unclear still how to respond to feedback. I see editors marking feedback such as "stuff", "Long Live Assange!", "Thanks Julian", "NO", "I strongly believe that Assange is innocent." as "RESOLVED". I would have thought these are (1) irrelevant, (2) not helpful for improving the article, or simply (3) an abuse of the feedback system (as such comments do not represent genuine feedback). --ELEKHH^T 22:50, 21 August 2012 (UTC)

I think we should just hide them. --Neo139 (talk) 20:17, 26 August 2012 (UTC)

Yeah hide. But only if we get consensus for it… :P benzband (talk) 20:35, 26 August 2012 (UTC)

Elekhh, Neo139 or Benzband what were you thinking in terms of wording for the hiding section? Callanecc (talk • contribs • logs) 12:12, 3 September 2012 (UTC)

Since there was no response here for a while, this discussion continued at Wikipedia talk:Article feedback#Dealing with non-actionable feedback. While "non-actionable" is accurate enough, "closed" seems to be a simple and short term, without the misleading potential of "resolved". --ELEKHH^T 13:23, 3 September 2012 (UTC)

A "close" or "no action needed" function sounds like a good idea if possible. benzband (talk) 16:55, 3 September 2012 (UTC)

I wish there was a way of excluding resolved feedback from the list. If it's resolved, why do we have to see it all the time? --99of9 (talk) 12:19, 3 September 2012 (UTC)

That's a comment for WT:AFT5, which is watched by people (namely :Okeyes (WMF)) who can do something about it. Callanecc (talk • contribs • logs) 12:40, 3 September 2012 (UTC)

Hiding duplicate posts

It seems to me that it would make sense to hide duplicate posts (and I've been doing this, in fact), simply as a method of cleaning up. I can naturally switch to resolving which ever of the posts has more information, but that seems not to clear out the stream. Comments? --Izno (talk) 00:37, 24 August 2012 (UTC)

Additionally, what about blank posts? --Izno (talk) 14:49, 25 August 2012 (UTC)

(I think) blank posts are automatically invisible, but not "hidden" in the sense that any user can see them by choosing the right setting in the pop-down menu. benzband (talk) 15:01, 25 August 2012 (UTC)

I've been seeing blank posts turn up in my watchlist feedback stream in the "all visible" filter... :/ --Izno (talk) 15:10, 25 August 2012 (UTC)

I'm in favor of hiding both (and have been). Callanecc (talk • contribs • logs) 15:20, 25 August 2012 (UTC)

Would it make sense to document that, perhaps as part of H6? --Izno (talk) 15:59, 25 August 2012 (UTC)

 Done as part of H6. Callanecc (talk • contribs • logs) 16:25, 25 August 2012 (UTC)

Hmm, exactly the same? I would imagine the minimum for hiding such a post would be the same editor, same page, and a substantially similar (80%+ the same) in content. Thoughts? --Izno (talk) 16:49, 25 August 2012 (UTC)

I wanted to try avoid different monitors applying it differently. And for it to have the purpose of bringing down the number of potentially useless (duplicated or blank) posts, and where the standard is very clear. However, now I think about it I think something like substantially similar (where the content is the same) would be better, as that allows for little things to be different, but the post to be pretty similar in other regards. How's that? Callanecc (talk • contribs • logs) 07:58, 26 August 2012 (UTC)

If by "content is the same" really means the content itself is the same but the wording is slightly different, or one where the wording is inclusive of slightly more "content" (but not one where the wording differs enough to be subtractive of content when one post is hidden), sure. At Special:ArticleFeedbackv5/Search engine optimization, we have the wording (the now hidden) "Actual i am find SEO information from web this artical very help full me because here is all information seo.", as opposed to (the not hidden) "Actually i am find some SEO information on web. It is very help full to me because here is available all information." Another case, at Special:ArticleFeedbackv5/T-shirt, "can you copy a t shirt" I would say is a duplicate (and hid it as such) of the earlier "can a t shirt with a picture on it becopied and reproduced". Would you agree to those hides?

I have concern now for hiding blank posts. They do include some information—not necessarily helpful, but not necessarily worthy of being hidden—in their use of the smiley faces. Does that actually make them worth blanking, or simply worthy of downvoting? --Izno (talk) 14:53, 26 August 2012 (UTC)

We had a discussion about this before, and the reason I changed my mind (from not hiding to hiding). In any case, the software will automatically hide blank feedback (it's just not working at the moment) so in that case it really would be non-contentious housekeeping. The reason for it was that at the moment the tool is active on around 10% of articles, so the amount of blank feedback (and so on) we get will increase substantially as the percentage is increased.

I agree regarding loosening the conditions on hiding duplicate posts, your comment brought me around. I've come up with the following, I know the wording is bit weird so feel free to edit it (that is, the bit on the box) so we can keep it together. Callanecc (talk • contribs • logs) 14:45, 29 August 2012 (UTC)

H5 Duplicate posts by the same user on the same article may be hidden if they provide the same or similar information as their duplicates. The most detailed or latest (monitor's discretion) post should be kept if it does not meet another of the criteria for hiding.

"Duplicate posts by the same user on the same article may be hidden if they provide the same or similar information as their duplicates. The most detailed or most recent post should be kept, at the monitor's discretion."

^ Edited copy. I've omitted "if it does not meet another of the criteria for hiding". I would suggest that both posts should be hidden for the primary reason rather than the fact that they are duplicates. I could lean either way though, so I'm not too worried. --Izno (talk) 18:19, 29 August 2012 (UTC)

Regarding monitor's discretion, I intended it be whether it was the most detailed or most recent which is to be kept, not whether any of them is kept at all. How about:

"Duplicate posts by the same user on the same article may be hidden if they provide the same or similar information as their duplicates. The most detailed or most recent post should be kept (it is the monitor's discretion as to which is kept). This criteron should be used secondarily to the others. That is, if H2 applies to one or some of the posts, use it first, then use this critera is there are still duplicates."

Callanecc (talk • contribs • logs) 05:02, 30 August 2012 (UTC)

Didn't mean to change the intention there! How about "Either the most detailed or most recent feedback should be kept, at the monitor's discretion." I think "posts" borders on jargon, so I removed it. Would there be a more appropriate word than 'post'?

How about "This criterion should be used only when another criterion would not cause the feedback to be hidden."? --Izno (talk) 17:14, 30 August 2012 (UTC)

Having it as a subordinate phrase afterward can always be read as the monitor's discretion if it is kept at all. What about "Monitors should determine whether the most detailed or the most recent feedback submission is to be left unhidden". Callanecc (talk • contribs • logs) 12:09, 3 September 2012 (UTC)

 Done I've added it as it's own criterion H6 and moved housekeeping to H7 and IAR to H8. The wording is fairlu close that what has been discussed above (I used feedback submission). Callanecc (talk • contribs • logs) 07:43, 6 September 2012 (UTC)

Missed this. The worry I have with your wording is that it doesn't mandate that at least one post be kept (which is apparently your concern with my wording as well). Additionally, I'm not sure that the example is necessary to explain the "if another criterion covers the posts", but I'm not too worried over that. I threw a bold edit or two in. Let me know what you think. --Izno (talk) 00:54, 7 September 2012 (UTC)

Looks good to me. Callanecc (talk • contribs • logs) 02:40, 7 September 2012 (UTC)

What to do with questions?

I've encountered a fair number of questions in the feedback on pharmaceutical drug articles ("Does clopidogrel reduce platelet count?", "Does fentanyl cause blindness?"). Apart from adding the queried information to the article (if appropriate), is there anything I can do to answer these? I normally answer such questions on talk pages (as long as this doesn't constitute medical advice, and even then I'd reply "please ask your physician"), but IPs who post feedback are probably unlikely to look there, and even may not see a post on their user talk page. Could we add a warning to the feedback form, on the lines of "If you expect an answer, please post on the talk page ['Add topic' link]"? Or are there any better ideas? --ἀνυπόδητος (talk) 18:02, 29 August 2012 (UTC)

I think this is an important issue and we need to get some resolution on it. I only just started looking at feedback, and I already see a lot of these kinds of questions and comments. They're neither helpful nor unhelpful, they're not really deserving of being featured or hidden, and they're not likely to result in changes to the article. Likewise, I see a lot of "I didn't find what I was looking for" feedback with comments indicating what they were looking for was something out of the article's scope, or for which there's debate on the article talk page. In software bug-reporting systems, this would generally be the kind of thing that would get a comment to explain why it's being flagged "Won't Fix". We need to be able to give that kind of response, and direct people to the article talk page for discussion, but I don't see a way to do it other than by marking it "resolved".

The proposed guideline for marking something resolved says Feedback should be marked as resolved if it has already been fulfilled (that is, it is already in the article). Feedback should also be marked as resolved if it includes a reasonable suggestion (which would normally meet the conditions for featuring) but is clearly not going to be fulfilled on policy grounds (please also state the policy as a comment when you mark it as resolved).

The proposed guideline for marking something featured says (emphasis mine) Any feedback which is reasonably actionable, may be useful to the community, is likely to improve the article and at some point is likely to be marked resolved either as done or as rejected (as opposed to unuseful), should be featured.

The latter guideline sort-of implies that the correct action is indeed to mark this kind of feedback as "resolved". However, the other guideline implies that rejected suggestions shouldn't be marked resolved unless the rejection is policy-based. Is that really what it means to say?

Regardless, the guidelines really need to explicitly address this situation. What do we do with suggestions which are reasonably rejected on non-policy grounds? —mjb (talk) 20:21, 12 September 2012 (UTC)

Ideally, we would have a won't fix button (probably "nicer" termed to avoid biting the newbies), which I have brought up on the main AFT talk page at the least. There won't be development on this for another couple months, last I heard. I probably need to crosspost that, or we need to start a list somewhere of desired functionality. --Izno (talk) 20:53, 12 September 2012 (UTC)

At the moment resolved should be anything which is not going to happen (so a question or comment about the subject which don't help with regard to improving the article) should be marked as resolved unless one of the criteria for hiding apply. The guidelines mean that any post which would reasonably assist in the improvement of the article are marked as featured. Following from there someone looks at the featured posts for the article and either decides to add the suggestion in (and marking as resolved with diff/reason) or decides not to make the change/follow through with suggestion (and marking as resolved wih reason). I agree that the guidelines could be clearer on this and I'll have a go at improving them in the next couple of days or so. Callanecc (talk • contribs • logs) 04:04, 13 September 2012 (UTC)

Changes on 12 September 2012

Two things I've done:

• Moved all reasons in H3 which were against people to H2 - to make them a little more different.
• Created a feedback protection section.

Any and all feedback and comments are welcome! Callanecc (talk • contribs • logs) 14:43, 12 September 2012 (UTC)

Feedback protection section

Just some background on the protection, this is from User:Yunshui (thank you) who had a look at the protection options for me:

"Basically it appears to be the same as regular protection: you can limit it by permissions (admin only, autoconfirmed only, logged-in only, everyone) and set an expiry time up to indefinite. Interestingly, it doesn't seem that the feedback protection level can exceed the page protection (so on a silverlocked page like Bieber's, you can't set the feedback protection to "admin-only")."

— User talk:Yunshui, User:Yunshui

Hope this helps everyone. Callanecc (talk • contribs • logs) 15:07, 12 September 2012 (UTC)